Academic Journal
SEAL: Semantic-Aware Contrastive Learning for scRNA-Seq Clustering.
| Title: | SEAL: Semantic-Aware Contrastive Learning for scRNA-Seq Clustering. |
|---|---|
| Authors: | Ye Y, Sun J, Peng L, Yang D, Xian J, Shen W, Liu C |
| Source: | IEEE transactions on computational biology and bioinformatics [IEEE Trans Comput Biol Bioinform] 2026 May-Jun; Vol. 23 (3), pp. 1136-1147. |
| Publication Type: | Journal Article |
| Language: | English |
| Journal Info: | Publisher: IEEE Country of Publication: United States NLM ID: 9919068173606676 Publication Model: Print Cited Medium: Internet ISSN: 2998-4165 (Electronic) Linking ISSN: 29984165 NLM ISO Abbreviation: IEEE Trans Comput Biol Bioinform Subsets: MEDLINE |
| Imprint Name(s): | Original Publication: [New York, New York] : IEEE, [2025]- |
| MeSH Terms: | Computational Biology*/methods , RNA-Seq*/methods , Clustering Algorithms* , Machine Learning* , Single-Cell Gene Expression Analysis*, Sequence Analysis, RNA/methods ; Animals ; Humans ; Cluster Analysis ; Semantics |
| Abstract: | The development of single-cell RNA sequencing (scRNA-seq) technology has enabled the exploration of biological processes at the cellular level. A critical task in scRNA-seq data analysis is the unsupervised clustering of cells to distinguish different cell types. While various clustering methods have been successfully developed for scRNA-seq data, they still face limitations, particularly in terms of unstable clustering performance. This is often due to their inability to fully capture the intrinsic properties of cells, especially in the presence of high dropout rates and noise in the data. In this work, we propose a SEmantic-Aware contrastive Learning (SEAL) approach for scRNA-seq clustering. Specifically, we randomly mask the gene expression of each cell to generate two different augmentations of the cell data, and then apply semantic-aware contrastive learning to capture semantically invariant representations across these augmentations by leveraging semantic information from generated pseudo-labels. Experimental results demonstrate that our method effectively learns biologically meaningful representations and accurately identifies cell types. |
| Entry Date(s): | Date Created: 20260303 Date Completed: 20260606 Latest Revision: 20260609 |
| Update Code: | 20260610 |
| DOI: | 10.1109/TCBBIO.2026.3669981 |
| PMID: | 41774662 |
| Database: | MEDLINE |
Be the first to leave a comment!